Home | Business News | Browse by Publication | S | Southwest Journal of Linguistics

On the development of pronominal clitics and affixes in Uto-Aztecan.

Article Excerpt
ABSTRACT. This paper documents the distribution of bound pronominal subjects and objects in Uto-Aztecan and addresses the likely historical development of both within the family. After critically reviewing a prominent proposal on the historical development of subject clitics in Uto-Aztecan (Steele 1977), I suggest an alternative proposal wherein the proto-language contained variable subject clitic placement (second position or pro-clitic to the verb). I then discuss the distribution of object clitics, and the relative placement of subject and object clitics within the verb complex of those Uto-Aztecan languages that allow (or require) both. Finally, I address the grammaticalization of Uto-Aztecan bound pronominals in light of the proposal by Baker (2001) bifurcating his POLYSYNTHESIS PARAMETER into two separate parameters: SUBJECT- VS. OBJECT POLYSYNTHESIS, arguing that the variability in linear pronominal ordering that we see within Uto-Aztecan resulted from different temporal FUSIONS of subject or object pronominals in the various languages (cf. Mithun 1991).

**********

1. INTRODUCTION. The Uto-Aztecan (UA) language family is one of the most widely distributed language families of the Americas, with languages of UA provenance being spoken in Western North and Middle America from the Great Basin to El Salvador. UA historical linguistics has a relatively long history, with the family being firmly established by the work of Sapir (1913 and 1915), but there is still much to be done to unravel the prehistory of these languages. Given the geographical range and time depth of this family, research into UA historical linguistics can shed light on a number of issues in the linguistic and cultural prehistory of a large section of Western North and Middle America. The purpose of the present paper is to address an important issue in comparative UA verb morphology and syntax: the historical development of subject and object pronominal clitics and affixes.

UA languages typically allow for (if they do not require) the marking of subject and/or object relations by means of pronominal clitics or affixes. However, the various UA languages differ with respect to where this marking occurs. This work documents the distribution of bound pronominal subjects and objects in UA to address the likely historical development of these clitics and affixes within the family.

The paper is organized as follows. Section 2 critically reviews the prominent proposal by Steele (1977) on the historical development of subject clitics in UA. This proposal suggests that UA prefixal subject pronominals originally developed from second position pronominals. Several empirical problems for such an analysis are presented and I argue for an alternative hypothesis whereby Proto-Uto-Aztecan (PUA) itself contained both second position and prefixal subject pronominals, either (or both) of which have been retained in different attested UA languages. Section 3 discusses the distribution of (direct) object clitics and affixes, and the relative placement of subject and object clitics in the verb complex of those languages that allow (or require) both. I argue that the variability in bound pronominal order seen in UA reflects the different chronological order of pronominal fusion to the verb complex, as has been proposed for other language families by Mithun (1991).

The concluding section addresses the grammaticalization of the UA pronominal patterns in light of Baker's recent (2001) proposal bifurcating the POLYSYNTHESIS PARAMETER into two separate parameters: SUBJECT- VS. OBJECT POLYSYNTHESIS. Baker suggests (but rejects) a possible need to split his polysynthesis macroparameter into two separate parameters, one for subject agreement and one for object agreement. Since the UA family has languages with obligatory agreement for the subject only and languages with obligatory agreement for the object only, in addition to those that require agreement for both, I argue that such a split is indeed necessary, if we are to accept the notion of a polysynthesis parameter at all.

Before proceeding I should note that I will not be making use of a possible distinction between the notions CLITIC and AFFIX. I simply assume the usual grammaticalization trajectory of bound pronominal elements, wherein free forms can develop into clitics and then affixes over time. The cognate elements that I will be discussing in UA are more relatively bound (i.e. affix-like) in some languages than in others, but since this is a language-specific distinction and the goal here is cross-linguistic generalizations based on comparative analysis I will not dwell on this matter in the pages to follow.

2. SUBJECT CLITICS IN UTO-AZTECAN. With regard to subject clitics, some UA languages have second position clitics, some have subject markers proclitic to the verb, some have both, and some have neither. Steele (1977) provides a useful summary of the distribution UA subject clitics, which I present as Table 1:

The Appendix gives a list of the UA languages and my sub-grouping, based on modification of Miller (1983). Since it is not especially important to the over-all claims of this paper I will not provide a discussion of UA family-internal sub-grouping here, other than to note that corresponding morphological patterns recurring across sub-groups are better evidence of a shared retention from Proto-Uto-Aztecan (PUA), all else being equal, than those that might occur among closely related languages within the same sub-group. An examination of the areal distribution of the languages in Table 1 leads to the conclusion that only second position clitics seem plausible as an areal phenomenon (as opposed to a retention from a common ancestor); for more on this topic see section 2.2.1.

Examples of second position subject clitics in Luiseno (Takic), Southern Paiute (Numic), and Tohono O'odham (formerly known as Papago, Tepiman) are presented in (1)-(3), respectively. Pronominal clitics and affixes are glossed as CP (for clitic pronominal), and subject clitics are marked in bold: (2)

(1) Luiseno hunwuti=pum $e?iwun bear:object=they(CP) are:shooting 'They are shooting a bear' (Steele 1977:539 [1]) (2) Southern Paiute qa[??]u=anga=ni n i ? imincuxwavaang?wain.iaanga NEG=object:CP=CP I shall:give:him:to:you 'I shall not give him to you' (Steele 1977:547 [17]) (3) Tohono O'odham ?aani=?an n i ok I=CP speaking 'I am/was speaking' (Steele 1977:547 [20])

Examples of proclitic or prefixal subject clitics from Classical Nahuatl (Aztecan), Cahuilla (Takic), and Kitanemuk (Takic) are presented in (4)-(6), respectively. Note that in each of these cases the proclitic/prefixal element is attached to a verb complex, as opposed to some other possible constituent within the sentence:

(4) Classical Nahuatl an-te[??]Laso?La (Steele 1977:539 [2]) you(CP)-love:us 'You love us' (5) Cahuilla [??]em [??]em-hi[??]iwe (Steele 1977: 564) we CP-are:going 'We are going' (6) Kitanemuk ni-kwara=mat pakwinin (Steele 1977: 564) CP-plaster=Aux with:mud 'I will plaster (cracks) with mud'

According to Steele, Yaqui and Tarahumara (both Taracahitan) can have both second position and proclitic subject pronominals, sometimes even in the same sentence, as shown in (7) and (8):

(7) Yaqui (3) kwarenta peso dyaryota=ne ne=koba iani inine forty peso daily=CP CP=earn now here 'Now I make forty pesos a day here' (Steele 1977:543 [9]) (8) Tarahumara [??]u mu [??]ika ke mu naki muhe ko ba WH CP WH NEG CP want you empatic emphatic 'Why don't you want it?' (Steele 1977:553 [30])

Also according to Steele, Tepecano (Tepiman) can have a clitic occupying both positions simultaneously, as in (9):

(9) Tepecano ndedos n=an=ahohoinda my:fingers introducer=CP=will:shake:them 'I will shake my fingers' (Steele 1977:543 [7])

Further examples of subject clitics in either position are also illustrated for Cupeno in (10):

(10) Cupeno a. ne?e=n yekwinqa I=CP be:scared 'I'm scared' b. [??]em=ep [?]e?-mamayew we=Aux...

View this article FREE - Now for a Limited Time, try Goliath Business News
Free for 3 Days!